Incidence and risk factors of symptomatic Petersen's hernias in bariatric and upper gastrointestinal surgery: a systematic review and meta-analysis.

PURPOSE

The aim of this study was to investigate the actual incidence of symptomatic Petersen's hernias (PH) as well as identify risk factors for their occurrence.

METHODS

Search was performed in Medline (via PubMed), Web of Science, and Cochrane library, using the keywords "Petersen Or Petersen's AND hernia" and "Internal hernia." Only studies of symptomatic PH were eligible. Fifty-three studies matched our criteria and were included. Risk of bias for each study was independently assessed using the checklist modification by Hoy et al. Analysis was performed using random-effects models, with subsequent subgroup analyses.

RESULTS

A total of 81,701 patients were included. Mean time interval from index operation to PH diagnosis was 17.8 months. Total small bowel obstruction (SBO) events at Petersen's site were 737 (0.7%). SBO incidence was significantly higher in patients without defect closure (1.2% vs 0.3%, p < 0.01), but was not significantly affected by anastomosis fashion (retrocolic 0.7% vs antecolic 0.8%, p = 0.99). SBO incidence was also not significantly affected by the surgical approach (laparoscopic = 0.7% vs open = 0.1%, p = 0.18). However, retrocolic anastomosis was found to be associated with marginally, but not significantly, increased SBO rate in patients with Petersen's space closure, compared with the antecolic anastomosis (p = 0.09).

CONCLUSION

PH development may occur after any gastric operation with gastrojejunal anastomosis. Contrary to anastomosis fashion and surgical approach, defect closure was demonstrated to significantly reduce SBO incidence. Limitations of this study may include the high heterogeneity and the possible publication bias across the included studies.